The key issue with using time as the key ‘measurable’ in a dosing application is that time is not volume – in other words we are not accurately measuring the volume of a particular substance but are approximating volume by time. To use an example, if one considers a large barrel of a particular chemical – for example chlorine – which is being dosed into a swimming pool at regular intervals to ensure the water remains fit for human use then the output of this barrel must be at the bottom of the barrel.
As the barrel reduces the pressure on the fluid will reduce. Whereas the first few doses will have been under additional pressure (as a result of gravitational pressure), the last few doses require careful pumping to be delivered. If time were used as the key variable by which the volume was determined, it is therefore quite likely that there may be a significant difference between the first dose out of the barrel (which would have been heavily gravity assisted) and the final dose (which would potentially have had gravity opposing it).
The result is a changing volume as the pressure changes.
NOTE: it is worth highlighting here that in applications where the input pressure on a pump is stable, time can be a very good measure provided all other factors remain constant (such as tube diameter, distance pumped etc.). However, in practice most applications have changing pressure of some sort so time will never be quite as accurate as other methods